928 
Proceedings of the Royal Society 
order to acquire the temperature of its surroundings ; when drawn 
up from depths under 10 fathoms, the mercury has not had time to 
recede from the index, and before it can he read evaporation from 
the wet bulb lowers the temperature and renders the reading 
inaccurate. 
Sir Kohert Christison’s cistern thermometer gave good results, 
hut it is too heavy an instrument to use continuously for many 
hours. 
For the first two sets of observations about to be described 
Kegretti and Zambra’s patent deep-sea thermometer with Magnaghi’s 
frame was used. The instrument is constructed so that after it has 
acquired the temperature of the water it may be turned over ; on 
this being done the mercury which has passed out of the bulb runs 
into the upper part of the tube which is now lowermost, and can 
be measured by means of a scale of degrees when the instrument is 
drawn up. Tt contains only one fluid — mercury ; and experiment 
has shown that three minutes are sufficient to enable it to acquire 
the temperature, although in our observations it was customary to 
allow four or five minutes as a minimum to make sure. The 
reversing arrangement of Magnaghi resembles a screw-propeller ; it 
is turned on drawing the instrument up through the water and 
releases the thermometer, which then turns over, as its pivots are 
below the centre of gravity and it never hangs quite perpen- 
dicularly. The objection to the method is that in shallow places 
the temperature at the bottom cannot be obtained, since the reversing 
screw will not act in much less than a fathom of water under the 
most favourable circumstances, and usually it requires considerably 
more. A modification of the thermometer was accordingly designed 
by Professor Chrystal and myself, and the necessary alterations were 
made in a most satisfactory manner by Mr Frazer of Lothian 
Street. 
The Scottish thermometer-frame, as the improved form has been 
named, differs from Magnaghi’s by the removal of the screw and fans 
from the pin that holds the thermometer in position, and the substi- 
tution in their place of a lever kept down by a spiral spring. The 
lever is forked at the outer end to allow the line to which the ther- 
mometer is attached to pass between. In order to prevent the line 
from slipping out of the fork, it is clasped by two thin springs 
